Microsoft's premium Copilot agents were confidently wrong

Microsoft is spending an insane amount of money on its AI features, building data centers and licensing large language models, with a goal to turn Windows and Microsoft 365 into an "agentic OS" that handles memos, presentations, meetings and routine tasks. Developers seem to be generally happy with the productivity gains they're seeing from tools like Claude Code and GitHub Copilot, but the agents working in the business sphere don't seem nearly as competent.

Over the past few weeks I tried the AI features in Microsoft 365 and Windows. Copilot shows occasional flashes of competence, but more often the results are a mishmash of misinformation, hallucinations and time-wasting dead ends. I paid the $10 to upgrade an unused account for a month and started with the Analyst agent, feeding it a household-spending spreadsheet and asking for design help.

It offered useful suggestions and even promised to "sketch a clean dashboard layout," then later claimed "I've created your modified workbook.

microsoft copilot, copilot agents, microsoft 365, windows, agentic os, llms, data centers, github copilot, claude code, hallucinations